Experimental Physics and
Industrial Control System

"Jeff Hill" <[email protected]> · Tue, 8 May 2001 15:09:29 -0600

Geoff,

At that location in the source code the CA client library is using
the close() system call to clean up the socket associated with the
virtual circuit for the IOC that disconnected.

Perhaps one of these might apply.

1) OSF1 does not return zero from socket_close() in all cases
when the request is successful. I would look closely at the
implementation of "base/src/include/os/osf1/osiSock.h". In 
particular, this macro might need to be modified. 

#define socket_close(S)         close(S)

2) There is general corruption of CA data structures.

3) Somehow close() is being called twice for the same 
socket. This appears to be less likely based on the location 
of the assert() call, the protections for this in the code,
and because this surely would have been detected on other OS.

Jeff

> -----Original Message-----
> From: Geoff Savage [mailto:[email protected]]
> Sent: Tuesday, May 08, 2001 2:26 PM
> To: [email protected]
> Subject: CA assert error on OSF1 version 5.1
> 
> 
> I was testing CA on OSF1 version 5.1 and received this error from CA
> when I rebooted the IOC on which the PV was located.  Has anyone tried
> this?
> 
> A call to "assert (status == 0)" failed in ../iocinf.c line 1472.
> The file "core" will be created in the current working directory.
> Please save this file and the text of this message in order to assist
> in diagnosing this problem.
> Please contact the author of this software or else send the text of
> this message to "[email protected]"
> This problem occurred in "@(#)Version R3.13.4 $2000/12/15 15:52:33$"
> Abort (core dumped)
> 
> Thanks
> 
> Geoff

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System